Darling, A., Rayman, M. P., Steer, C., Golding, J., Lanham-New, S., & Bath,S. (2017). Association between maternal vitamin D status in pregnancy andneurodevelopmental outcomes in childhood: results from the AvonLongitudinal Study of Parents and Children (ALSPAC). British Journal ofNutrition, 117(12), 1682-1692. https://doi.org/10.1017/S0007114517001398

Peer reviewed version

License (if available):Unspecified

Link to published version (if available):10.1017/S0007114517001398

Link to publication record in Explore Bristol ResearchPDF-document

This is the author accepted manuscript (AAM). The final published version (version of record) is available onlinevia Cambridge University Press at https://www.cambridge.org/core/journals/british-journal-of-nutrition/article/association-between-maternal-vitamin-d-status-in-pregnancy-and-neurodevelopmental-outcomes-in-childhood-results-from-the-avon-longitudinal-study-of-parents-and-children-alspac/F4DE023F16B77075E6B0A526EFFD426B . Please refer to any applicable terms of use of the publisher.

University of Bristol - Explore Bristol ResearchGeneral rights

This document is made available in accordance with publisher policies. Please cite only the publishedversion using the reference above. Full terms of use are available:http://www.bristol.ac.uk/pure/about/ebr-terms

1

1

Association between maternal vitamin D status in pregnancy and neurodevelopmental 1

outcomes in childhood; results from the Avon Longitudinal Study of Parents and Children 2

(ALSPAC) 3

4

Andrea L Darling1, Margaret. P Rayman1, Colin D Steer2, Jean Golding2, Susan A Lanham-New1, 5

and Sarah C Bath1* 6

7

1Department of Nutritional Sciences, School of Biosciences and Medicine, Faculty of Health and 8

Medical Sciences, University of Surrey, Guildford, UK (ALD, MPR, SAL-N, SCB), 2Centre for 9

Child and Adolescent Health, School of Social and Community Medicine, University of Bristol, 10

Bristol, UK (CDS, JG) 11

12

*Corresponding Author: Dr Sarah Bath, Department of Nutritional Sciences, School of 13

Biosciences and Medicine, Faculty of Health and Medical Sciences, University of Surrey, 14

Guildford, GU2 7XH. Telephone +044 (01483) 683631. Email: s.bath@surrey.ac.uk 15

16

Short title: Maternal vitamin D and offspring development 17

18

Keywords: prenatal vitamin D, 25–hydroxy–vitamin D, motor development, social development, 19

IQ and reading ability, ALSPAC 20

21

22

2

2

Abstract 23

24

Seafood intake in pregnancy has been positively associated with childhood cognitive outcomes 25

which could potentially relate to the high vitamin-D content of oily fish. However, whether higher 26

maternal vitamin D status [serum 25-hydroxy-vitamin D, 25(OH)D] in pregnancy is associated with 27

a reduced risk of offspring suboptimal neurodevelopmental outcomes is unclear. A total of 7065 28

mother-child pairs were studied from the Avon Longitudinal Study of Parents and Children 29

(ALSPAC) cohort who had data for both serum total 25(OH)D concentration in pregnancy and at 30

least one measure of offspring neurodevelopment (pre-school development at 6–42 months; 31

“Strengths and Difficulties Questionnaire” scores at 7 years; IQ at 8 years; reading ability at 9 32

years). After adjustment for confounders, children of vitamin-D deficient mothers (< 50.0 nmol/L) 33

were more likely to have scores in the lowest quartile for gross motor development at 30 months 34

(OR 1.20 95% CI 1.03, 1.40), fine motor development at 30 months (OR 1.23 95% CI 1.05, 1.44), 35

and social development at 42 months (OR 1.20 95% CI 1.01, 1.41) than vitamin-D sufficient 36

mothers (≥ 50.0 nmol/L). No associations were found with neurodevelopmental outcomes, 37

including IQ, measured at older ages. However, our results suggest that deficient maternal vitamin 38

D status in pregnancy may have adverse effects on some measures of motor and social development 39

in children under 4 years. Prevention of vitamin D deficiency may be important for preventing 40

suboptimal development in the first 4 years of life. 41

42

43

3

3

Introduction 44

The consumption of fish, or nutrients present in fish, by pregnant women has been linked to 45

neurocognitive development in their children. In observational studies, maternal intake of fish or 46

seafood in pregnancy has been positively associated with cognitive scores in the offspring(1; 2; 3; 4), 47

while children whose mothers had eaten oily fish in early pregnancy had a reduced risk of 48

hyperactivity than those whose mothers did not eat oily fish(3). While these studies tended to 49

interpret these associations as effects of of long-chain omega-3 fatty acids, they might also be 50

explained by the fact that oily fish is the best dietary source of vitamin D. Though the action of 51

sunlight on the skin is the predominant contributor to vitamin D status, dietary vitamin D can play 52

an important role in determining status, as measured by the vitamin D metabolite, 25-53

hydroxyvitamin D [25(OH)D]1, in serum or plasma(5). Dietary sources of vitamin D (especially 54

oily fish) are particularly important during the winter months when endogenous production of 55

vitamin D status is limited. 56

57

It is biologically plausible that vitamin D status in pregnant mothers may affect child 58

neurocognitive development as vitamin D receptors are present in the brain(6) and maternal vitamin 59

D deficiency is known to be associated with abnormal brain development in the young rat(7). In the 60

period from birth to weaning in rats, there appears to be a window during which maternal vitamin D 61

status affects offspring brain development(8) and these developmental changes may not occur if 62

vitamin D is withheld until weaning(9). Furthermore, vitamin D deficiency in late gestation can lead 63

to impaired brain function in adult rats(8). Due to differences between rat and human developmental 64

physiology, the extent to which these findings would apply to humans remains unclear. 65

66

Few human studies have assessed the relationship between maternal vitamin D status and 67

neurodevelopmental outcomes. The results of the five published observational studies that exist are 68

inconsistent(10; 11; 12; 13; 14). Indeed, this fact was recently highlighted in the report from Public Health 69

England on Vitamin D and Health from the Scientific Advisory Committee for Nutrition (SACN) 70

(15). 71

72

To address this lack of consistent evidence with respect to the association between maternal vitamin 73

D status and cognitive-developmental outcomes in the offspring, we analysed data from the Avon 74

Longitudinal Study of Parents and Children (ALSPAC) cohort. Our a priori hypothesis was that 75

poorer maternal vitamin D status, as measured by serum 25(OH)D, would be associated with 76

4

4

increased probability of suboptimal cognitive or behavioural development scores in childhood of 6 77

months to 9 years. 78

Subjects and Methods 79

80

Study Design and Participants 81

Details of ALSPAC methods have been detailed previously (16). In brief, all pregnant women living 82

in the former Avon area in southwest England, who had an expected delivery date between April 1st 83

1991 and December 31st 1992 were eligible for inclusion. A total of 14,541 women were recruited, 84

and there were 13,617 mother-child pairs with singleton offspring alive at one year. The ALSPAC 85

study website contains details of all the data that are available through a fully searchable data 86

dictionary (http://www.bris.ac.uk/alspac/). Our study sample consisted of mother-child pairs that 87

had both a serum 25(OH)D measure in pregnancy and at least one neurodevelopmental outcome of 88

interest from 6 months to 9 years (Figure 1). A range of outcomes was explored, including motor 89

development, communication and social skills, behaviour, cognition and reading ability. 90

91

Outcomes 92

The ALSPAC pre-school development tests, which were based on questionnaires completed by the 93

mother when the child was between 6 and 42 months of age, provided scores for four domains: fine 94

motor, gross motor, social development, and communication (details published previously(1)). The 95

Strengths and Difficulties Questionnaire (SDQ)(17) was completed by mothers when the child was 96

81 months of age and was used to assess behavioural development. Intelligence Quotient (IQ) at age 97

8 years had been assessed in the ALSPAC clinic using the abbreviated form of the Wechsler 98

Intelligence Scale for Children, as previously described (1). Reading ability (accuracy, 99

comprehension and speed) was assessed at age 9 years by trained psychologists using the Neale 100

Analysis of Reading Ability(18) and by asking children to read real words to derive a reading score. 101

Further details of these outcomes are available in the Supplementary File. 102

103

Maternal vitamin D status 104

Although 25(OH)D has lower biological activity than the active vitamin D hormone, 1,25-105

dihydroxyvitamin D [1,25(OH)2D], serum/plasma 25(OH)D is widely regarded as the most 106

reliable marker of vitamin D status(19). Total maternal serum 25(OH)D concentration (including 107

both vitamin D2 and vitamin D3) in ALSPAC mothers had been measured in a previous study by 108

high-performance liquid chromatography and tandem mass-spectrometry, in accordance with 109

5

5

Vitamin D External Quality Assessment Scheme (DEQAS) requirements; full details have been 110

published previously(20), including details of inter-assay coefficients of variation(21). 111

112

Statistical analysis 113

The women with vitamin D measurements were compared to the remaining ALSPAC women. We 114

compared categorical variables with χ2 tests and continuous variables with independent t-tests. We 115

used median (IQR, Inter-quartile Range) to describe maternal vitamin D status. Our main analysis 116

dichotomised women as deficient or sufficient using 25(OH)D concentration ≤ 50.0 nmol/L as the 117

cut-off for vitamin D deficiency, as in previous ALSPAC work(20). We did additional 118

supplementary analyses by dividing women into three categories (< 25.0, 25.0–49.9 and ≥ 50.0 119

nmol/L) to explore the dose-response relationship. 120

121

We used logistic regression to examine the relationship between maternal vitamin D status in 122

pregnancy and odds of suboptimal development with the women in the vitamin-D-sufficient group 123

(> 50.0 nmol/L) as the reference category. We did not input missing confounder or outcome data 124

with replacement values. We defined suboptimal development as scores in the lowest quartile for all 125

subscales of early development, IQ and reading ability, as in previous ALSPAC research (1; 22). For 126

the SDQ, suboptimal behaviour was defined according to published cut-offs (for both the individual 127

scales and overall score) that indicate borderline/abnormal behaviour (17) (see Supplementary File 128

Study Outcomes). Model predictors were assessed for potential multicollinearity. For our final 129

model, variance inflation factor ranged from 1.02 to 2.2 (accordingly tolerance ranged from 0.5-130

0.99) depending on the variable. 131

132

As vitamin D status and childhood cognitive and behavioural development are affected by a range 133

of factors(23; 24), we included potential confounders in our analysis. The confounders chosen were 134

based on previous ALSPAC findings(1; 22) and were from questionnaire and clinic-based data (Table 135

1). We included ten categorical and two continuous variables. The two continuous variables were 136

maternal age (years), and maternal body mass index (BMI, Kg/m2). As there is a well-established 137

relationship between BMI and 25(OH)D concentration(25), maternal BMI was included in the model, 138

even though it was not statistically associated with 25(OH)D in this dataset (Table 1). 139

140

The ten categorical variables comprised three groups: (i) child factors [gender and breastfeeding 141

(none or some)], (ii) maternal factors [ethnicity (white or non-white), tobacco use in the first 142

trimester (smoker or non-smoker), parity (zero, one or more) and oily fish intake in pregnancy 143

6

6

(never/rarely or once a fortnight or more)], and (iii) markers of socio–economic development 144

[maternal education (low = less than O-level or equivalent; medium = O-level, and high = greater 145

than O-level), home ownership (mortgaged/owned, privately rented or housing association/council 146

rented/other), maternal social class based on her occupation (non-manual and manual) and 147

crowding in the home (≤ one person or > one person per room)]. We also included two variables to 148

control for variation in the vitamin D measurement: gestation (week) and season of sample 149

collection [spring (March, April, and May), summer (June, July, and August) autumn (September, 150

October, and November), and winter (December, January and February)]. While it is unlikely that 151

the age of the child at assessment would be confounded by maternal vitamin D status, outcomes 152

were adjusted for child age at the 6-month measurement, owing to the strong association between 153

age and outcomes at this early life stage. 154

155

We used three models to adjust the analysis for potential confounders. As 25(OH)D measurements 156

spanned pregnancy, and as gestational week is associated with vitamin D status(26), we do not 157

present unadjusted data; our minimally adjusted model (Model 1) included gestational week of 158

25(OH)D measurement. Model 2 built on Model 1 by including nine confounders associated with 159

both vitamin D status (Table 1) and cognitive development (parity, tobacco smoking, housing 160

status, crowding, maternal age, BMI, education, ethnic group, and social class) and two child 161

factors (gender and breastfeeding). Model 3 included Model 2 confounders plus two variables (oily 162

fish intake and season of vitamin D measurement) that could affect maternal vitamin D status 163

though including these may represent an over-control. 164

165

We used simulations to assess the impact of multiple comparisons. We generated 5000 datasets 166

where 25(OH)D measurements were randomly permutated across valid observations with these 167

data. As a consequence, all analyses maintained the same number of observations and, with all other 168

data unchanged, the correlations between outcomes and confounders were preserved. The analyses 169

were based upon Model 3. The effect of randomisation was to generate a set of results under the 170

null hypothesis to which our set of observed results could be compared. A composite score across 171

the 27 outcomes was based upon the sum of P values. These were modified to one-sided tests to 172

allow results in the same direction to contribute consistently to the score, whether statistically 173

significant or not. P values in the tables are not corrected for multiple comparisons. 174

175

176

177

7

7

Sensitivity analysis 178

We conducted analyses with two additional confounders (added to Model 3) that might be on the 179

causal pathway: preterm birth (< 37 weeks or ≥ 37 weeks) and birth weight (< 2500 g or ≥ 2500 g). 180

We also explored the effect of including maternal iodine status in the first trimester [sufficient (≥ 181

150 μg/g) or deficient (< 150 μg/g)] as we have previously shown that this is associated with child 182

cognition in the ALSPAC cohort(22). As just 787 women also had a measure of iodine status in the 183

first trimester, we used a simplified model (total of 13 confounders) to ensure that the model would 184

converge (we dropped ethnicity and crowding in the home as a result of low numbers in the 185

categories of those variables). 186

187

As there is ongoing controversy in the published literature with respect to the definition of vitamin 188

D deficiency(27), we conducted sensitivity analyses using a wide range of vitamin D status, namely 189

< 25.0 and < 75.0 nmol/L as cut–offs (Supplementary Tables 3 and 4). Assumptions concerning 190

statistical significance were based on interpretation of confidence intervals, rather than P values, 191

wherever possible, and multiple testing was assessed as described above. Analyses were conducted 192

using the Statistical Package for Social Sciences (version 21·0; SPSS, Inc., Chicago, USA). 193

194

Ethics 195

The ALSPAC study was conducted according to the guidelines laid down in the Declaration of 196

Helsinki. All procedures involving human subjects were approved by the ALSPAC Ethics and Law 197

Committee and the Local Research Ethics Committees. Written informed consent was obtained 198

from participants (or from their parent/guardian if under 18 years old). 199

200

Role of the funding source 201

The funding bodies did not have a role in the study design, data collection, data analysis, data 202

interpretation, or writing of the report. The corresponding author had full access to all the study data 203

used and final responsibility to submit for publication. 204

205

Results 206

Compared with the remainder of the ALSPAC cohort (defined as mother-singleton child pairs from 207

the core sample surviving to one year), the mother-child pairs in this study were more likely to be 208

older, of white ethnicity, with markers of higher socio-economic status [e.g. a higher proportion of 209

breast-feeding mothers, higher educational attainment and social class, and a lower proportion of 210

smokers (Supplementary Table 1)]. However, some of the actual differences were small (e.g. 211

8

8

maternal age 28.3 (4.8) vs. 27.7 (4.7) years). The median (IQR) 25(OH)D concentration for all 7065 212

women with a child that had at least one relevant outcome was 61.3 (42.9 – 84.7) nmol/L, with 213

4.4% having < 25.0 nmol/L, 34.6% having < 50.0 nmol/L and 65.7% having < 75.0 nmol/L. 214

215

The median (IQR) gestational week of vitamin D measurement (available for 7064 women) was 216

29.6 (12.7, 33.3) weeks, with 26.1% in the first trimester (≤ 13 weeks), 11.8% in the second 217

trimester (14 – 27 weeks) and 62.1% in the third trimester (≥ 28 weeks). The median (IQR) 218

25(OH)D measurement was 54.9 (40.1 - 72.5) nmol/L in the first trimester, 59.3 (38.6 - 84.2) 219

nmol/L in the second trimester and 65.3 (45.2 - 90.4) nmol/L in the third trimester. Table 1 shows 220

the confounders associated with maternal vitamin D status using the 50 nmol/L cut-off. Women 221

with 25(OH)D concentration ≥ 50.0 nmol/L were more likely to be white, older, and have markers 222

of higher socio-economic status (for example education, home ownership and reduced smoking and 223

crowding). 224

225

Results of logistic regression models using the cut-off value for serum 25(OH)D of <50.0 nmol/L to 226

define deficiency are shown in Table 2. In the minimally adjusted analysis (Model 1), the only 227

outcomes associated with vitamin D status were verbal IQ at 8 years and words read per minute at 228

age 9 (Table 2). However, after adjustment for potential confounders, the effect on IQ and reading 229

was attenuated and the only outcomes that remained statistically significant were gross- and fine-230

motor development at 30 months and social development at 42 months. With further adjustment for 231

oily-fish intake and season (Model 3), the association between maternal vitamin D status and gross-232

motor development also became significant at 18 months, while remaining associated with gross-233

motor and fine-motor development at 30 months and social development at 42 months (Table 2). 234

Children born to mothers with 25(OH)D ≤ 50.0 nmol/L were more likely to have scores in the 235

bottom quartile for these variables. 236

237

For the ALSPAC pre-school development assessments, when the serum 25(OH)D of < 50.0 nmol/L 238

group was divided into < 25.0 and 25.0 – 49.9 nmol/L, there was evidence of a statistically 239

significant trend to decreasing risk of suboptimal development with higher maternal 25(OH)D 240

concentration for gross-motor skills at 18 (P=0.02) and 30 months (P=0.008), fine-motor skills at 30 241

months (P=0.01) and social development at 42 months (P=0.02), after adjustment for all 12 242

confounders in Model 3 (Table 3). The effect sizes were larger for odds of suboptimal development 243

in children of mothers in the serum 25(OH)D < 25.0 nmol/L group, than for the serum 25(OH)D of 244

9

9

25.0 – 49.9 nmol/L group (with the ≥ 50.0 nmol/L group as the comparison group) for all outcomes 245

except fine-motor development at 18 months and social development at 30 months. 246

247

The interaction between gestational week of 25(OH)D measurement and the vitamin D variable 248

(i.e. deficient vs. sufficient status) was significant for only two of 27 outcomes: fine-motor skills at 249

30 months and performance IQ (Table 4). However, when the analysis was restricted to the 250

ALSPAC pre-school development assessments and was split into early (≤22 weeks) and late 251

gestation (> 22 weeks), the results suggested that the effect of deficient vs. sufficient vitamin D 252

status on the majority of tests was greater in the second half of gestation. The effect sizes were 253

generally larger in the second half of gestation and results were significant (Table 4) for gross motor 254

development at 18 months (Odds Ratio (OR) 0.97, 95% CI 0.76, 1.23 vs. OR 1.31, 95% 1.08, 1.58) 255

and 30 months (OR 1.07, 95% CI 0.84,1.38 vs OR 1.28, 95% CI 1.05,1.57), fine motor 256

development at 30 months (0.99, 95%CI 0.76,1.29 vs OR. 1.37, 95% CI 1.12,1.67) and social 257

development at 42 months (OR 1.07, 95% CI 0.82,1.41 vs. OR 1.28, 95% CI 1.03,1.58). There were 258

no significant associations in either half of gestation for other neurodevelopmental outcomes, 259

including the SDQ, IQ or reading ability (Table 4). 260

261

Multiple comparisons 262

While only 4 results in Table 2 were nominally significant at the 5% level, it was noted that 25 of 263

the 27 results in Model 3 showed a detrimental effect for low vitamin D status. Such a result would 264

be highly significant (p<0.0001) if the outcomes were independent. In practice, outcomes were 265

correlated with an average r = 0.12 (range –0.03 to 0.69). The impact of these correlations was 266

assessed using simulations. The scores from the 5000 simulated datasets had a mean (SD) of 13.52 267

(2.78). This compared to an expected mean (SD) of 13.5 (1.5) if all the outcomes had been 268

independent. The observed results had a score of 6.93 suggesting an empirical two-tail P value of 269

0.016. Sequential analyses by removing those outcomes with the strongest association from the 270

simulated scores suggested that three outcomes (gross and fine motor development at 30 months 271

and social development at 42 months) had robust associations with the other 24 outcomes having 272

associations consistent with chance (p=0.051). 273

274

We also explored defining the score based upon the logit transformation, ln(p/(1-p)). Using this 275

definition, the score more closely approximated to a normal distribution. However this did not 276

change the conclusions. 277

278

10

10

Sensitivity analysis 279

When we added the variables, preterm birth and birth weight, to Model 3, the results were 280

fundamentally unchanged (Supplementary Table 2), though the effect of maternal vitamin D status 281

on gross motor development at 18 months and social development at 42 months was no longer 282

statistically significant. 283

284

The addition of suboptimal iodine-to-creatinine ratio in the first trimester to Model 3 resulted in 285

considerable sample attrition given the low number of women with iodine measurements (n=787) 286

(Supplementary Table 2). Though the effect sizes were larger than previously, the associations 287

between maternal vitamin D and gross motor development at 18 and 30 months and social 288

development at 42 months were no longer significant, though they remained significant for fine 289

motor development at 18 (OR 1.50, 95% CI 1.02, 2.23) and 30 months (OR 1.61, 95% CI 1.06, 290

2.46). 291

292

We explored whether dichotomising women according to different 25(OH)D cut-offs (25.0 or 75.0 293

nmol/L) changed the results (Supplementary Tables 3 and 4), bearing in mind the lower relative 294

statistical power that results when the cut-off leads to unequal numbers in each group (the 50.0 295

nmol/L cut-off was close to the median 25(OH)D concentration of 54.9 nmol/L). When using the 296

25.0 nmol/L cut-off, the only outcome associated with vitamin D deficiency in the fully adjusted 297

model was gross motor development at 30 months (OR 1.43 95% CI 1.01-2.02); results approached 298

statistical significance for other outcomes (e.g. social development at 42 months, OR 1.40 95% CI 299

0.97-2.02; Supplementary Table 3). Using a cut-off of 75.0 nmol/L to define deficiency resulted in 300

null associations with the ALSPAC pre-school development assessments, behaviour and cognitive 301

tests, but was associated with higher odds of sub-optimal reading accuracy at 9 years (OR 1.26 95% 302

CI 1.01, 1.57); however, this may be a chance finding as reading accuracy was not associated with 303

vitamin D in any other analyses (Tables 2, 3 and 4 and Supplementary Tables 2 and 3). 304

305

Discussion 306

After adjustment for potential confounders, children born to vitamin-D deficient mothers (serum 307

25(OH)D of <50.0 nmol/L) were more likely to have sub-optimal gross-motor skills at 30 months, 308

sub-optimal fine-motor skills at 30 months and sub-optimal social development scores at 42 months 309

than were children born to sufficient mothers (≥50.0 nmol/L). Although the effect sizes were 310

relatively small, we consider that the findings were biologically meaningful. Interestingly, no 311

11

11

associations were found between maternal vitamin D status and other outcomes (IQ, reading 312

ability). 313

314

These results suggest that the vitamin D content of seafood might explain some of the beneficial 315

effects of maternal seafood consumption seen previously in ALSPAC, at least for fine-motor skills 316

at 30 months and social skills at 42 months(1). The classification of maternal seafood consumption 317

by Hibbeln et al.(1) included white fish and shellfish which are not good sources of dietary vitamin 318

D, therefore, we would not expect vitamin D intake to account totally for their findings. 319

Furthermore, our results cannot explain previous associations found in ALSPAC between maternal 320

seafood consumption and IQ(1) or between maternal iodine status and IQ and reading ability(22). 321

322

Our findings on fine- and gross-motor skills support previous non-ALSPAC-based research that 323

found a positive association between maternal vitamin D status and infant psychomotor 324

development(11). Although we did not specifically measure scholastic achievement, the lack of an 325

association between maternal vitamin D status and either reading ability or IQ in our study 326

reinforces the findings of a previous study that found no relationship between maternal 25(OH)D 327

status and offspring scholastic achievement(10). While a US study found a relationship between 328

maternal vitamin D status and offspring IQ, the effect estimates were very small and there was very 329

little indication of an association between maternal blood 25(OH)D and cognitive development, 330

achievement, or behaviour between 8 months and 7 years of age(12). 331

332

Our findings suggest that some specific aspects of early neurocognitive development may be 333

suboptimal if maternal prenatal vitamin D is deficient (i.e. serum 25(OH)D of < 50.0 nmol/L) in 334

pregnancy. The biological mechanism underpinning this association in humans is not fully 335

understood, but the ubiquitous presence of the vitamin D receptor (VDR) and the hydroxylase 336

enzymes controlling vitamin D metabolism in a wide variety of areas of the human brain(6), as well 337

as neurological developmental mechanisms previously identified in studies of vitamin D deficiency 338

in pregnant rats may be relevant(7; 9; 28; 29). These include enlarged brain ventricles, thinner 339

neocortex(29), and more mitotic cells in the brain(29), suggesting a less differentiated phenotype(28). 340

The active form of vitamin D [1,25(OH)2D], may also affect the development of the brain by 341

influencing the production of cytokines(30), affecting neurotransmission(31) and synaptic plasticity(31) 342

which is likely to affect learning processes(32) and therefore neurocognitive development. 343

1,25(OH)2D likely affects dopamine activity in the brain owing to the presence of the vitamin D 344

receptor (VDR) in brain areas responsive to dopamine(33). Ventral midbrain dopaminergic neurones 345

12

12

are known to play a key role in the modulation of motor behaviour(34). It is therefore feasible that 346

1,25(OH)2D may affect motor development via its effects on the dopaminergic system. Other 347

potential mechanisms may relate to an association between maternal 25(OH)D status and fetal 348

growth retardation (e.g. reduced fetal head size) which is associated with later developmental 349

disabilities(35). A recent study in the Generation R cohort in the Netherlands found an association 350

between lower maternal 25(OH)D status at 20 weeks gestation and smaller fetal-head circumference 351

in the third trimester(36), suggesting that poorer maternal 25(OH)D status may predispose children to 352

developmental delay via effects on intra-uterine growth restriction. 353

354

When we assessed the impact of gestational age on our results for outcomes that were significantly 355

associated with vitamin D in the main analyses, we found that the effect sizes were generally 356

greater when vitamin D was measured in the second half (> 22 weeks) than in the first half (≤ 22 357

weeks) of pregnancy. There is a small amount of evidence in rats that re-introduction of vitamin D 358

after birth, but before end of weaning, can rescue normal brain development(28); that time period 359

correspond to the third trimester in humans, suggesting a potential crucial window for vitamin D in 360

brain development. However, all interpretations in our analysis of gestational timing need to be 361

interpreted in light of the fact that we only had one measurement of maternal vitamin D status for 362

each woman and so we cannot draw clear conclusions on the effects of gestational timing of vitamin 363

D deficiency. Furthermore, we cannot be sure that our observed effects are confined to the 364

gestational week that the 25(OH)D measurement was made, as some individuals may have 365

persistent pattern of vitamin D status that extends into later pregnancy or infancy. 366

367

When the women were split into three groups [serum 25(OH)D of <25.0, 25.0 – 49.9 and ≥ 50.0 368

nmol/L], adverse outcomes were present in the offspring of mothers with insufficient status (serum 369

25(OH)D < 50nmol/L) as well as those with severe deficiency (serum 25(OH)D < 25nmol/L). 370

However, there was a trend to larger effect sizes in the more deficient < 25.0 nmol/L group than in 371

the 25.0 – 49.9 nmol/L group; the relatively small sample size in the < 25.0 nmol/L group explains 372

the wider confidence intervals seen for this cut-off. The outcomes that were significantly associated 373

with vitamin D when women were dichotomised on the basis of a cut-off of 50.0 nmol/L were not 374

significant when the cut-off was increased to 75.0 nmol/L. These findings support a vitamin D 375

status cut-off for optimal child outcomes closer to 50.0 nmol/L than to 75.0 nmol/L. 376

377

As the women in the ALSPAC study were recruited over 20 years ago, we compared their vitamin 378

D status to more recent measurements in UK women to assess the current relevance of our findings. 379

13

13

As 25(OH)D status does not differ between pregnant and non–pregnant women(15) we looked at 380

nationally representative data in UK women from the recent National Diet and Nutrition Survey 381

(NDNS). In the latest report (sampling 2008/9 – 2011/12), 21.7% of women of 19–64 years had a 382

plasma 25(OH)D concentration below 25 nmol/L(37), a higher percentage than the 4.4% of women 383

in ALSPAC. Other studies(38; 39), including those in pregnancy, suggest that many UK women are 384

vitamin D deficient. Currently, the UK National Institute for Health and Care Excellence (NICE) 385

recommends that pregnant women should take a supplement of 10 µg (400 IU) of vitamin D per 386

day(40). However use of vitamin D supplements in pregnancy is low, with a recent survey (2005–387

2009) finding that only 1.4% of UK pregnant women had taken a vitamin D supplement(41). Our 388

findings give further evidence that public-health campaigns should address the vitamin D status of 389

UK pregnant women, and encourage compliance with the 10 µg/d recommendation(40). 390

391

Strengths and Limitations 392

Although our study has several strengths, including the large sample size, there are also limitations. 393

Firstly each woman had only one measure of maternal vitamin D status in pregnancy which may not 394

have reflected status over the whole of pregnancy. In addition, the range of vitamin D status in the 395

ALSPAC women was limited, with approximately one third (34.6%) having a 25(OH)D 396

concentration less than 50.0 nmol/L and only a small proportion having a 25(OH)D concentration 397

less than 25.0 nmol/L (4.4%). Moreover, ALSPAC only has a relatively small number of women 398

from ethnic-minority backgrounds (just 2% of this study sample), who are known to be at particular 399

risk of having low 25(OH)D concentrations(42), suggesting that the results may differ in populations 400

with a larger number of ethnic-minority individuals. Finally, we were not able to control for the 401

association between infant vitamin D status and neurocognitive function as we had no measures of 402

vitamin D status in infancy. Infant vitamin D status may partly explain some of the association seen 403

in this paper between maternal vitamin D status and infant neurodevelopment. 404

405

In conclusion, we found that maternal vitamin D status in pregnancy was associated with a number 406

of adverse neurocognitive developmental variables in early childhood, albeit with a small, but 407

nonetheless important, effect size. There is a need for replication of this work in other settings to 408

confirm these results, but the public-health implications of these findings are nevertheless 409

potentially important. Further study is now urgently required, particularly in population groups that 410

are more severely vitamin D deficient such as dark-skinned ethnic-minority women37 who may 411

show a wider range and greater severity of sub-optimal neurocognitive outcomes. 412

413

14

14

Acknowledgments 414

We are extremely grateful to all the families who took part in this study, the midwives for their help 415

in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and 416

laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists and 417

nurses. 418

419

Sources of support 420

This work was supported by the following funding bodies: The UK Medical Research Council and 421

Wellcome Trust (grant number 102215/2/13/2) and the University of Bristol currently provide core 422

support for ALSPAC. The work was also supported by a UK Medical Research Council Population 423

Health Scientist Fellowship (S.C.B., grant number MR/K02132X/1). The UK Medical Research 424

Council provided funds to ALSPAC for completion of the 25(OH)D assays used in this paper (grant 425

number G0701603). 426

427

Conflict of Interest 428

SLN is Research Director of D3Tex Ltd which holds the UK Patent (Gulf Cooperation Council 429

Patent pending) on the use of ultraviolet-B (UVB) transparent material for vitamin D deficiency 430

prevention. All other authors declare that they have no conflicts of interest. 431

432

Authors’ Contributions to the Manuscript 433

ALD, SCB and JG designed the current research project. SCB and ALD conducted the statistical 434

analyses with statistical advice from CDS, MPR and JG. MPR, JG, CDS and SLN revised the 435

paper and made suggestions on the content. ALD and SCB wrote the paper. SCB has primary 436

responsibility for final content. 437

438

439

15

15

References 440

1. Hibbeln JR, Davis JM, Steer C et al. (2007) Maternal seafood consumption in pregnancy and 441 neurodevelopmental outcomes in childhood (ALSPAC study): an observational cohort study. 442

Lancet 369, 578-585. 443 2. Daniels JL, Longnecker MP, Rowland AS et al. (2004) Fish intake during pregnancy and early 444

cognitive development of offspring. Epidemiology 15, 394-402. 445 3. Gale CR, Robinson SM, Godfrey KM et al. (2008) Oily fish intake during pregnancy - 446

association with lower hyperactivity but not with higher full-scale IQ in offspring. J Child Psychol 447 Psyc 49, 1061-1068. 448

4. Julvez J, Mendez M, Fernandez-Barres S et al. (2016) Maternal Consumption of Seafood in 449 Pregnancy and Child Neuropsychological Development: A Longitudinal Study Based on a 450

Population With High Consumption Levels. Am J Epidemiol 183, 169-182. 451 5. Brock K, Huang WY, Fraser DR et al. (2010) Low vitamin D status is associated with physical 452

inactivity, obesity and low vitamin D intake in a large US sample of healthy middle-aged men and 453 women. J Steroid Biochem 121, 462-466. 454

6. Eyles DW, Smith S, Kinobe R et al. (2005) Distribution of the vitamin D receptor and 1 alpha-455 hydroxylase in human brain. J Chem Neuroanat 29, 21-30. 456

7. Eyles DW, Feron F, Cui X et al. (2009) Developmental vitamin D deficiency causes abnormal 457 brain development. Psychoneuroendocrinology 34 Suppl 1, S247-257. 458

8. O'Loan J, Eyles DW, Kesby J et al. (2007) Vitamin D deficiency during various stages of 459 pregnancy in the rat; its impact on development and behaviour in adult offspring. 460 Psychoneuroendocrinology 32, 227-234. 461

9. Feron F, Burne TH, Brown J et al. (2005) Developmental Vitamin D3 deficiency alters the adult 462 rat brain. Brain Res Bull 65, 141-148. 463

10. Strom M, Halldorsson TI, Hansen S et al. (2014) Vitamin D measured in maternal serum and 464 offspring neurodevelopmental outcomes: a prospective study with long-term follow-up. Ann Nutr 465

Metab 64, 254-261. 466 11. Morales E, Guxens M, Llop S et al. (2012) Circulating 25-hydroxyvitamin D3 in pregnancy and 467

infant neuropsychological development. Pediatrics 130, e913-920. 468 12. Keim SA, Bodnar LM, Klebanoff MA (2014) Maternal and cord blood 25(OH)-vitamin D 469

concentrations in relation to child development and behaviour. Paediatr Perinat Epidemiol 28, 434-470 444. 471

13. Whitehouse AJ, Holt BJ, Serralha M et al. (2012) Maternal serum vitamin D levels during 472 pregnancy and offspring neurocognitive development. Pediatrics 129, 485-493. 473

14. Hanieh S, Ha TT, Simpson JA et al. (2014) Maternal vitamin D status and infant outcomes in 474 rural Vietnam: a prospective cohort study. PloS one 9, e99005. 475

15. SACN (2016) Vitamin D and Health. https://www.gov.uk/government/publications/sacn-476 vitamin-d-and-health-report (accessed 29th March 2017) 477

16. Boyd A, Golding J, Macleod J et al. (2013) Cohort Profile: the 'children of the 90s'--the index 478 offspring of the Avon Longitudinal Study of Parents and Children. Int J Epidemiol 42, 111-127. 479

17. Goodman R (1997) The Strengths and Difficulties Questionnaire: a research note. J Child 480 Psychol Psychiatry 38, 581-586. 481

18. Neale M (1997) Neale Analysis of Reading Ability- Revised. Windsor: NFER-Nelson. 482 19. Seamans KM, Cashman KD (2009) Existing and potentially novel functional markers of 483

vitamin D status: a systematic review. Am J Clin Nutr 89, 1997s-2008s. 484 20. Lawlor DA, Wills AK, Fraser A et al. (2013) Association of maternal vitamin D status during 485

pregnancy with bone-mineral content in offspring: a prospective cohort study. Lancet 381, 2176-486 2183. 487

21. Sullivan S, Wills A, Lawlor D et al. (2013) Prenatal vitamin D status and risk of psychotic 488 experiences at age 18years-a longitudinal birth cohort. Schizophr Res 148, 87-92. 489

16

16

22. Bath SC, Steer CD, Golding J et al. (2013) Effect of inadequate iodine status in UK pregnant 490 women on cognitive outcomes in their children: results from the Avon Longitudinal Study of 491

Parents and Children (ALSPAC). Lancet 382, 331-337. 492 23. Tolppanen AM, Fraser A, Fraser WD et al. (2012) Risk factors for variation in 25-493

hydroxyvitamin D(3) and D(2) concentrations and vitamin D deficiency in children. J Clin 494 Endocrinol Metab 97, 1202-1210. 495

24. Schoon I, Jones E, Cheng H et al. (2012) Family hardship, family instability, and cognitive 496 development. J Epidemiol Community Health 66, 716-722. 497

25. Samuel L, Borrell LN (2013) The effect of body mass index on optimal vitamin D status in U.S. 498 adults: the National Health and Nutrition Examination Survey 2001-2006. Ann Epidemiol 23, 409-499

414. 500 26. Kuoppala T, Tuimala R, Parviainen M et al. (1986) Serum levels of vitamin D metabolites, 501

calcium, phosphorus, magnesium and alkaline phosphatase in Finnish women throughout 502 pregnancy and in cord serum at delivery. Hum Nutr Clin Nutr 40, 287-293. 503

27. Heaney RP (2013) Health is better at serum 25(OH)D above 30 ng/mL. J Steroid Biochem 136, 504 224-228. 505

28. Cui X, Gooch H, Groves NJ et al. (2015) Vitamin D and the brain: key questions for future 506 research. J Steroid Biochem 148, 305-309. 507

29. Eyles D, Brown J, Mackay-Sim A et al. (2003) Vitamin D3 and brain development. 508 Neuroscience 118, 641-653. 509

30. van Etten E, Mathieu C (2005) Immunoregulation by 1,25-dihydroxyvitamin D3: basic 510 concepts. J Steroid Biochem 97, 93-101. 511 31. Almeras L, Eyles D, Benech P et al. (2007) Developmental vitamin D deficiency alters brain 512

protein expression in the adult rat: implications for neuropsychiatric disorders. Proteomics 7, 769-513 780. 514

32. Becker A, Eyles DW, McGrath JJ et al. (2005) Transient prenatal vitamin D deficiency is 515 associated with subtle alterations in learning and memory functions in adult rats. Behav Brain Res 516

161, 306-312. 517 33. Kesby JP, Eyles DW, Burne TH et al. (2011) The effects of vitamin D on brain development 518

and adult brain function. Mol Cell Endocrinol 347, 121-127. 519 34. Van den Heuvel DMA, Pasterkamp RJ (2008) Getting connected in the dopamine system. Prog 520

Neurobiol 85, 75-93. 521 35. Watemberg N, Silver S, Harel S et al. (2002) Significance of microcephaly among children with 522

developmental disabilities. J Child Neurol 17, 117-122. 523 36. Miliku K, Vinkhuyzen A, Blanken LM et al. (2016) Maternal vitamin D concentrations during 524

pregnancy, fetal growth patterns, and risks of adverse birth outcomes. The American journal of 525 clinical nutrition 103, 1514-1522. 526

37. Bates B, Lennox A, Prentice A et al. (2014) National Diet and Nutrition Survey: Results from 527 Years 1, 2, 3 and 4 (combined) of the Rolling Programme (2008/2009 – 2011/2012). London: 528

Public Health England. 529 38. Macdonald HM, Mavroeidi A, Fraser WD et al. (2011) Sunlight and dietary contributions to the 530

seasonal vitamin D status of cohorts of healthy postmenopausal women living at northerly latitudes: 531 a major cause for concern? Osteoporosis Int 22, 2461-2472. 532

39. Javaid MK, Crozier SR, Harvey NC et al. (2006) Maternal vitamin D status during pregnancy 533 and childhood bone mass at age 9 years: a longitudinal study. Lancet 367, 36-43. 534

40. NICE (2014) Vitamin D: increasing supplement use in at-risk groups [PH56]. 535 http://www.nice.org.uk/guidance/ph56 (accessed 15th November 2015 536

41. Oliver EM, Grimshaw KEC, Schoemaker AA et al. (2014) Dietary Habits and Supplement Use 537 in Relation to National Pregnancy Recommendations: Data from the EuroPrevall Birth Cohort. 538

Matern Child Hlth J 18, 2408-2425. 539

17

17

42. Cashman KD, Dowling KG, Skrabakova Z et al. (2016) Vitamin D deficiency in Europe: 540 pandemic? The American journal of clinical nutrition 103, 1033-1044. 541

542

18

18

Table 1 Relationship between confounders and maternal Vitamin D status Confounder Maternal vitamin D status

< 50.0 nmol/L ≥ 50.0 nmol/L

Mean SD n Mean SD n p value†

Age of mother (yrs) 27.7 4.8 2443 28.6 4.7 4622 < 0·0001

BMI of mother (Kg/m2) 23.0 4.0 2126 22.9 3.6 4095 0.43

Gestation of vitamin D measure

(weeks)

23.4 10.9 2771 25.7 10.3 5174 < 0·0001

% n % n p value ‡

Breastfeeding

Some 33·0% 1738 67·0% 3526 < 0·0001

None 38·8% 553 61·2% 874

Crowding in the home

< one person per room 33·9% 2140 66·1% 4170 < 0·0001

One or more per room 43.6% 176 56·4% 228

Education of mother

Low 37·5% 716 62·5% 1195 < 0·0001

Medium 33·4% 792 66·6% 1577

High 31·5% 755 68·5% 1643

Ethnicity of mother

White 33·3% 2171 66·7% 4344 < 0·0001

Non–white 60·6% 83 39·4% 54

Gender of child

Male 34·3% 1266 65·7% 2421 0·67

Female 34·8% 1177 65·2% 2201

Housing status

Owned/mortgaged 32·8% 1705 67·2% 3487 < 0·0001

Other rented 36·6% 150 63·4% 260

Council rented 41·0% 491 59·0% 708

Iodine–to–creatinine ratio in 1st trimester

<150 μg/g (deficient) 33.5% 186 66.5% 374 0.94

≥150 μg/g (sufficient) 33.2% 76 66.8% 151

Oily fish intake in pregnancy (/week)

Never/rarely 37·7% 1038 62·3% 1718 < 0·0001

Once per fortnight or more 31·3% 1191 68·7% 2617

Parity

Zero 37·0% 1125 63·0% 1914 < 0·0001

One or more 31·9% 1179 68·1% 2516

Season of vitamin D measure

Spring 48.8% 980 51.2% 1027 < 0·0001

Summer 15.2% 268 84.8% 1491

Autumn 22.4% 363 77.6% 1257

Winter 49.5% 831 50.5% 847

Smoking in 1st trimester

No tobacco 31·7% 1652 68·3% 3567 < 0·0001

Smoked tobacco 42·5% 689 57·5% 932

Social class of mother

Manual 36·6% 383 63·4% 664 0·01

Non–manual 32·5% 1447 67·5% 3008

† p value from independent t-test.

‡p value for χ2 test.

19

19

Table 2 Odds of suboptimal outcomes according to maternal vitamin D status (< 50.0 vs ≥ 50.0 nmol/L), minimally and fully adjusted for

potential confounders Model 1† Model 2‡ Model 3§

Age OR (95% CI) p value n OR (95% CI) p value n OR (95% CI) p value n

ALSPAC

pre–school

development

assessments

Gross Motor

Skills

6 mo ‖ 0.96 (0.84, 1.09) 0.49 6242 1.01 (0.86, 1.18) 0.92 4383 0.96 (0.81, 1.13) 0.59 4380

18 mo 0.98 (0.87, 1.10) 0.74 6269 1.10 (0.96, 1.27) 0.18 4385 1.17 (1.01, 1.36) 0.04 4383

30 mo 1.02 (0.91, 1.16) 0.71 5843 1.16 (1.00, 1.34) 0.05 4135 1.20 (1.03, 1.40) 0.02 4133

42 mo 0.99 (0.87, 1.13) 0.89 5695 1.04 (0.89, 1.22) 0.60 4073 1.09 (0.92, 1.28) 0.31 4070

Fine Motor

Skills

6 mo ‖ 0.93 (0.82, 1.05) 0.24 5880 1.07 (0.92, 1.25) 0.39 4141 1.06 (0.91, 1.25) 0.47 4139

18 mo 1.07 (0.96, 1.21) 0.24 6268 1.03 (0.90, 1.19) 0.65 4383 1.09 (0.94, 1.27) 0.26 4381

30 mo 1.09 (0.96, 1.23) 0.18 5854 1.20 (1.04, 1.40) 0.02 4138 1.23 (1.05, 1.44) 0.01 4136

42 mo 1.04 (0.92, 1.19) 0.51 5692 1.11 (0.95, 1.31) 0.19 4071 1.16 (0.98, 1.37) 0.08 4068

Social

Development

6 mo ‖ 0.96 (0.84, 1.09) 0.52 6010 1.02 (0.87, 1.19) 0.81 4209 1.00 (0.85, 1.18) 0.98 4207

18 mo 1.01 (0.89, 1.15) 0.86 6268 1.10 (0.94, 1.28) 0.22 4383 1.14 (0.97, 1.34) 0.11 4381

30 mo 0.97 (0.86, 1.10) 0.64 5843 1.11 (0.95, 1.30) 0.18 4129 1.07 (0.91, 1.27) 0.42 4127

42 mo 1.04 (0.92, 1.18) 0.54 5689 1.19 (1.02, 1.39) 0.03 4069 1.20 (1.01, 1.41) 0.04 4066

Communication 6 mo ‖ 0.99 (0.85, 1.15) 0.90 6100 0.99 (0.83, 1.20) 0.95 4285 0.99 (0.81, 1.20) 0.90 4283 18 mo 0.99 (0.87, 1.12) 0.85 6279 1.11 (0.96, 1.29) 0.17 4390 1.12 (0.95, 1.31) 0.18 4388

Behaviour Prosocial 7 yr 0.92 (0.75, 1.13) 0.40 4791 0.97 (0.75, 1.24) 0.78 3513 1.00 (0.77, 1.31) 0.98 3511

Peer problems 7 yr 1.05 (0.88, 1.25) 0.58 4785 1.03 (0.83, 1.27) 0.80 3510 1.05 (0.83, 1.31) 0.70 3508

Hyperactivity 7 yr 1.06 (0.91, 1.24) 0.47 4780 1.04 (0.86, 1.26) 0.68 3513 1.04 (0.85, 1.26) 0.74 3511 Emotional 7 yr 1.17 (0.98, 1.41) 0.09 4785 1.14 (0.92, 1.42) 0.23 3511 1.20 (0.95, 1.51) 0.12 3509

Conduct 7 yr 1.13 (0.99, 1.30) 0.08 4790 1.05 (0.88, 1.24) 0.60 3514 1.06 (0.89, 1.27) 0.50 3512

Total Score 7 yr 1.08 (089, 1.32) 0.42 4777 1.13 (0.89, 1.44) 0.31 3510 1.24 (0.96, 1.60) 0.09 3508

Cognition Verbal IQ 8 yr 1.19 (1.02, 1.39) 0.03 3997 1.08 (0.89, 1.31) 0.47 2952 1.00 (0.82, 1.23) 0.98 2950

Performance IQ 8 yr 1.06 (0.91, 1.24) 0.43 3990 0.99 (0.82, 1.20) 0.92 2945 1.00 (0.82, 1.23) 0.98 2943

Total IQ 8 yr 1.16 (1.00, 1.35) 0.06 3978 1.02 (0.84, 1.24) 0.82 2938 1.01 (0.82, 1.24) 0.93 2936

Reading

ability

Words per min 9 yr 1.17 (1.00, 1.36) 0.05 3794 1.14 (0.94, 1.39) 0.18 2763 1.15 (0.94, 1.42) 0.17 2761 Accuracy 9 yr 1.16 (0.99, 1.35) 0.07 3802 1.04 (0.85, 1.28) 0.69 2767 1.03 (0.83, 1.27) 0.80 2765

Comprehension 9 yr 1.11 (0.95, 1.30) 0.18 3802 1.02 (0.83, 1.25) 0.87 2767 1.04 (0.84, 1.29) 0.73 2765

Reading Score 9 yr 1.10 (0.95, 1.27) 0.22 4125 1.06 (0.88, 1.27) 0.54 3028 1.04 (0.86, 1.26) 0.69 3026 mo, month; OR, odds ratio; n, number of subjects; yr, years. Suboptimal outcome defined as scores in the bottom quartile for ALSPAC pre–school development assessments, cognition, and reading ability. Published cut-offs(17)

were used for behaviour: Prosocial (≤5; 9·8%), Peer problems (≥3; 13·5%), hyperactivity (≥6; 18·7%), emotional symptoms (≥4; 12·2%), conduct problems (≥3; 24·3%), and total score (≥14; 10·5%). Maternal vitamin D status

>50.0 nmol/L was the reference group. †Model 1 adjusted for gestational week of vitamin D measurement; ‡Model 2: gestational week of vitamin D measurement plus additional 11 variables: maternal age, maternal BMI,

maternal ethnic group, maternal education, maternal social class, parity, tobacco smoking in 1st trimester, home ownership status, crowding index, child gender, breastfeeding; §Model 3: additionally adjusted for oily fish and

season of vitamin D measurement; ‖ age of child at development test included in all models.

20

20

Table 3 Odds of suboptimal outcomes in offspring according to maternal vitamin D status when the < 50.0 nmol/L group is split into < 25.0

and 25.0 – 49.9 nmol/L and each group is compared to ≥ 50.0 nmol/L (adjusted model 3). Maternal vitamin D status (nmol/L)

< 25.0 vs. ≥ 50.0 25.0 – 49.9 vs. ≥ 50.0 Trend

OR (95% CI) n OR (95% CI) n p value n

ALSPAC pre–

school

development

assessments

Gross Motor

Skills

6 mo† 1.30 (0.90, 1.88) 169 0.92 (0.77, 1.09) 1279 0.88 4380

18 mo 1.40 (1.00, 1.96) 178 1.14 (0.98, 1.33) 1270 0.02 4383

30 mo 1.52 (1.07, 2.17) 163 1.17 (0.99, 1.37) 1213 0.008 4133

42 mo 1.24 (0.85, 1.82) 159 1.07 (0.90, 1.27) 1191 0.23 4070

Fine Motor

Skills

6 mo† 1.24 (0.85, 1.80) 167 1.04 (0.88, 1.24) 1213 0.32 4139

18 mo 1.03 (0.72, 1.47) 177 1.10 (0.94, 1.29) 1269 0.36 4381

30 mo 1.30 (0.91, 1.88) 163 1.22 (1.04, 1.44) 1214 0.01 4136

42 mo 1.31 (0.89, 1.92) 158 1.14 (0.96, 1.36) 1191 0.06 4068

Social Development

6 mo† 1.02 (0.70, 1.50) 170 1.00 (0.84, 1.19) 1216 0.95 4207 18 mo 1.28 (0.88, 1.85) 177 1.12 (0.95, 1.33) 1269 0.08 4381

30 mo 0.91 (0.61, 1.36) 163 1.09 (0.92, 1.30) 1212 0.66 4127

42 mo 1.49 (1.02, 2.18) 158 1.16 (0.98, 1.38) 1190 0.02 4066

Communication 6 mo† 1.41 (0.93, 2.14) 167 0.94 (0.77, 1.16) 1237 0.59 4283

18 mo 1.31 (0.92, 1.88) 179 1.09 (0.93, 1.29) 1272 0.11 4388

Behaviour Prosocial 7 yr 1.11 (0.59, 2.09) 124 0.99 (0.75, 1.30) 1003 0.89 3511

Peer problems 7 yr 0.97 (0.56, 1.67) 124 1.05 (0.84, 1.33) 1002 0.80 3508

Hyperactivity 7 yr 0.63 (0.37, 1.08) 124 1.09 (0.89, 1.33) 1002 0.70 3511

Emotional 7 yr 0.80 (0.43, 1.49) 124 1.25 (0.99, 1.57) 1002 0.34 3509

Conduct 7 yr 0.80 (0.50, 1.27) 124 1.10 (0.91, 1.32) 1003 0.88 3512

Total Score 7 yr 0.68 (0.33, 1.39) 124 1.31 (1.02, 1.70) 1001 0.37 3508

Cognition Verbal IQ 8 yr 1.07 (0.67, 1.73) 103 0.99 (0.80, 1.23) 839 0.90 2950

Performance IQ 8 yr 1.40 (0.89, 2.20) 104 0.96 (0.78, 1.18) 837 0.56 2943

Total IQ 8 yr 1.37 (0.87, 2.17) 103 0.97 (0.78, 1.20) 834 0.54 2936

Reading ability Words per min 9 yr 1.11 (0.68, 1.80) 101 1.16 (0.94, 1.43) 797 0.23 2761 Accuracy 9 yr 1.14 (0.70, 1.87) 101 1.02 (0.81, 1.27) 799 0.69 2765

Comprehension 9 yr 1.01 (0.61, 1.66) 101 1.04 (0.84, 1.30) 799 0.78 2765

Reading Score 9 yr 0.91 (0.57, 1.45) 108 1.06 (0.87, 1.29) 872 0.88 3026 mo, month; OR, odds ratio; n, number of subjects; yr, years. Suboptimal outcome defined as scores in the bottom quartile for ALSPAC pre–school development assessments, cognition, and reading ability. Published cut-offs(17) were used for behaviour: Prosocial (≤5; 9·8%), Peer problems (≥3; 13·5%), hyperactivity (≥6; 18·7%), emotional symptoms (≥4; 12·2%), conduct problems (≥3; 24·3%), and total score (≥14; 10·5%). Maternal vitamin D status ≥ 50.0 nmol/L was the reference group. †age of child at development test included in all models.

21

21

Table 4 Odds of suboptimal outcomes in offspring by maternal vitamin D status (< 50.0 vs ≥ 50.0 nmol/L) according to whether maternal

vitamin D was measured in the first or second half of gestation (Adjusted Model 3)

First half of gestation (≤ 22 weeks) Second half of gestation (> 22 weeks) P value for interaction*

OR (95% CI) P value n OR (95% CI) P value n

ALSPAC

pre–school

development

assessments

Gross Motor Skills

6 mo† 0.92 (0. 70, 1.22) 0.56 1500 0.98 (0.79, 1.21) 0.84 2880 0.21 18 mo 0.97 (0.76, 1.23) 0.78 1522 1.31 (1.08, 1.58) 0.005 2861 0.13

30 mo 1.07 (0.84, 1.38) 0.58 1435 1.28 (1.05, 1.57) 0.02 2698 0.79

42 mo 1.03 (0.79, 1.34) 0.85 1422 1.10 (0.89, 1.36) 0.37 2648 0.72

Fine Motor

Skills

6 mo† 1.09 (0.83, 1.44) 0.52 1436 1.03 (0.83, 1.27) 0.80 2703 0.25

18 mo 1.05 (0.82, 1.36) 0.69 1522 1.10 (0.90, 1.33) 0.35 2859 0.46

30 mo 0.99 (0.76, 1.29) 0.95 1436 1.37 (1.12, 1.67) 0.002 2700 0.05

42 mo 1.03 (0.78, 1.37) 0.83 1420 1.24 (1.00, 1.53) 0.05 2648 0.37

Social

Development

6 mo† 0.88 (0.66, 1.16) 0.37 1453 1.11 (0.90, 1.38) 0.32 2754 0.90

18 mo 1.23 (0.95, 1.60) 0.12 1522 1.07 (0.87, 1.32) 0.51 2859 0.11

30 mo 0.96 (0.74, 1.26) 0.79 1431 1.13 (0.91, 1.40) 0.28 2696 0.36 42 mo 1.07 (0.82, 1.41) 0.62 1420 1.28 (1.03, 1.58) 0.02 2646 0.26

Communication 6 mo† 0.90 (0.65, 1.23) 0.50 1468 1.04 (0.81, 1.34) 0.75 2815 0.37

18 mo 1.27 (0.98, 1.65) 0.07 1524 1.04 (0.85, 1.28) 0.71 2864 0.17

Behaviour Prosocial‡ 7 yr 0.75 (0.48, 1.17) 0.21 1216 1.15 (0.83, 1.61) 0.40 2301 0.10

Peer problems 7 yr 1.14 (0.78, 1.66) 0.49 1210 0.97 (0.73, 1.30) 0.86 2298 0.55

Hyperactivity 7 yr 0.95 (0.68, 1.33) 0.75 1213 1.10 (0.86, 1.41) 0.46 2298 0.31

Emotional‡ 7 yr 1.25 (0.87, 1.80) 0.23 1214 1.17 (0.87, 1.58) 0.29 2301 0.71

Conduct 7 yr 1.13 (0.84, 1.52) 0.42 1212 1.04 (0.82, 1.31) 0.74 2300 0.76

Total Score‡ 7 yr 1.20 (0.79, 1.82) 0.40 1214 1.24 (0.90, 1.71) 0.18 2300 0.79

Cognition Verbal IQ 8 yr 1.09 (0.77, 1.55) 0.64 1025 0.93 (0.72, 1.21) 0.60 1925 0.20

Performance IQ 8 yr 1.15 (0.83, 1.59) 0.42 1017 0.89 (0.68, 1.16) 0.38 1926 0.03

Total IQ 8 yr 1.18 (0.84, 1.66) 0.33 1015 0.90 (0.69, 1.17) 0.43 1921 0.13

Reading

ability

Words per min 9 yr 1.41 (1.00, 1.97) 0.05 936 1.00 (0.77, 1.31) 0.98 1825 0.20

Accuracy 9 yr 1.31 (0.92, 1.87) 0.13 938 0.87 (0.66, 1.14) 0.32 1827 0.06

Comprehension 9 yr 1.09 (0.77, 1.55) 0.62 938 0.98 (0.75, 1.29) 0.89 1827 0.31

Reading Score 9 yr 1.30 (0.94, 1.78) 0.11 1060 0.91 (0.71, 1.16) 0.44 1966 0.20 mo, month; OR, odds ratio; n, number of subjects; yr, years. Suboptimal outcome defined as scores in the bottom quartile for ALSPAC pre–school development assessments, cognition, and reading ability. Published cut-offs(17)

were used for behaviour: Prosocial (≤5; 9·8%), Peer problems (≥3; 13·5%), hyperactivity (≥6; 18·7%), emotional symptoms (≥4; 12·2%), conduct problems (≥3; 24·3%), and total score (≥14; 10·5%). Maternal vitamin D status

≥ 50.0 nmol/L was the reference group and Model 3 was used (without gestational week of vitamin D assessment as this was used to split analyses). *interaction between vitamin D (deficient/sufficient) and gestational week of

sample (continuous variable); †age of child at development test included in all models; ‡ethnicity removed as model would not converge.

22

22

Legends for Figures

Figure 1: Flow of participants

23

23

Figure 1